Biopsychosocial Compromise
ags
13-08-2020
Dataset Match
We used the Treatment Outcomes Profile to join users of the C1 (n= 110,403; users=85,603) that were evaluated with the Treatment Outcome Profile (TOP) in the same date of Admission to entries in C1 datasets.
###casos duplicados en fecha de aplicación
#1 1 97779
#2 2 2483
#3 3 149
#4 4 21
#5 5 3
#no hay casos que tengan más de una fecha de admisión Y UNA BD MÁS RECIENTE
#CONS_TOP_df_dup_ENE_2020_prev9%>% dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ap_top)))%>% group_by(concat) %>% dplyr::mutate(n_hash_fech_ap=n(),n_dis_ano=n_distinct(ano_bd))%>% dplyr::filter(n_hash_fech_ap>1,n_dis_ano>1)
invisible(
CONS_TOP_df_dup_ENE_2020_prev9%>% dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ap_top)))%>% group_by(concat) %>% dplyr::mutate(n_hash_fech_ap=n(),n_dis_ano=n_distinct(ano_bd))%>% dplyr::filter(n_hash_fech_ap>1,n_dis_ano>1)
)
CONS_TOP_df_dup_ENE_2020_prev9_match<-
CONS_TOP_df_dup_ENE_2020_prev9%>% dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ap_top)))%>% group_by(concat) %>% dplyr::mutate(n_hash_fech_ap=n(),n_dis_ano=n_distinct(ano_bd))%>% dplyr::filter(n_hash_fech_ap==1)%>%
dplyr::select(hash_key,concat,total_oh,dosis_oh,total_thc,dosis_thc,total_pbc,dosis_pbc,total_coc,dosis_coc,total_bzd,dosis_bzd,total_otra,dosis_otra,hurto,robo,venta_drogas,rina,total_vif,otro,total_transgresion,salud_psicologica,total_trabajo,total_educacion,salud_fisica,lugar_vivir,vivienda,calidad_vida,etapa_del_tratamiento)
CONS_C1_df_dup_JUL_2020_match_top<-
CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(compromiso_biopsicosocial), etapa_del_tratamiento=="Inicio Tratamiento")%>%
dplyr::select(-hash_key_TOP)%>%
dplyr::filter_at(vars(total_oh,dosis_oh,total_thc,dosis_thc,total_pbc,dosis_pbc,total_coc,dosis_coc,total_bzd,dosis_bzd,total_otra,dosis_otra,hurto,robo,venta_drogas,rina,total_vif,otro,total_transgresion,salud_psicologica,total_trabajo,total_educacion,salud_fisica,lugar_vivir,vivienda,calidad_vida,etapa_del_tratamiento), all_vars(!is.na(.)))
users_top_only_one_app_match<-CONS_TOP_df_dup_ENE_2020_prev9%>% dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ap_top)))%>% group_by(concat) %>% dplyr::mutate(n_hash_fech_ap=n(),n_dis_ano=n_distinct(ano_bd))%>% dplyr::filter(n_hash_fech_ap==1)%>% dplyr::ungroup()%>% dplyr::distinct(hash_key)%>% nrow()
#c("total_oh","dosis_oh","total_thc","dosis_thc","total_pbc","dosis_pbc","total_coc","dosis_coc","total_bzd","dosis_bzd","total_otra","dosis_otra","hurto","robo","venta_drogas","rina","total_vif","otro","total_transgresion","salud_psicologica","total_trabajo","total_educacion","salud_fisica","lugar_vivir","vivienda","calidad_vida")
# etapa_del_tratamiento n percent valid_percent
# Egreso 4 3.470114e-05 0.0005719188
# Inicio Tratamiento 6948 6.027587e-02 0.9934229339
# Seguimiento 12 meses 0 0.000000e+00 0.0000000000
# Seguimiento 15 meses 0 0.000000e+00 0.0000000000
# Seguimiento 3 meses 34 2.949597e-04 0.0048613097
# Seguimiento 6 meses 8 6.940227e-05 0.0011438376
# Seguimiento 9 meses 0 0.000000e+00 0.0000000000
# <NA> 108276 9.393251e-01 NAOf the total entries in TOP dataset (n= 110,403), we did not considered 2,845 entries in TOP dataset had more than one application in the same date (3%). We selected 6,390 cases that had no differences between the date of application and the date of admission to treatment. Additionally, these applications must had been administered at admission (excluding the 0,6% of applications that were administered at other stage of treatment). Finally, we discarded cases with null values in the variables of interest (230 entries).
Explore variables
First, values of doses over 50 were discarded and declared missing, due to the great amount of outliers.
We defined one outcome variable to indicate early discharge (abandono_temp) that should vary depending on biopsychosocial compromise. We also collapsed variables related to transgression to norms (transg_norm) into a continuous variable, by counting actions committed in the last 4 weeks. Additionally, we added the variable salud_mean to get the mean of the scores in physical health, quality of life and psychological health. To capture the substance use behaviors, we defined the average pattern of drug use in alcohol, marijuana, cocaine paste base, cocaine sedatives and tranquilizers, and other substances, in terms of scores of quantity in the last four weeks (total_mean). Equally, we defined a score of the average dose of the mean reported by users on each substance. At last, we created the mean of scores of days of attendance to jobs or educational institutions in the last four weeks (trabajo_edu_mean).
Furthermore, we were interested in co-occurrent physical and psychiatric conditions among substance users. Consequently, we generated a variable called dg_cie_10, composed of users that at least had been diagnosed with a psychiatric condition or professionals had been studying a possible diagnosis. Also, we included dg_trs_fisico, which categorizes between users without a diagnosed physical condition (0), in study (1), and those diagnosed with a physical condition (2).
To capture profiles of consumption, we generated the following variables: oh identifies cases that used alcohol at least one day in the last four weeks. This same criteria was applied to marijuana (thc), cocaine paste base (pbc), cocaine (coc), benzodiazepines (bzd), and other drugs (otra). Users that reported at least two substances according to the aforementioned rule were categorized as polydrug users (polycons).
Lastly, the mean of the scores in many variables were dichotomized between users that had less than the 50% of users, compared to people that showed more scores than the 50% of the population. These were defined as Higher Frequency of Days using substances in the last four weeks (mayor_total), Higher Average Reported Daily Dose of Substances (mayor_dosis), Lower Health-Related Self-Reports (menor_salud), and Lower Days of Attendance to Work and Educational Centers in the last four weeks (menor_trabajo_edu). In a similar manner, transg_norm_bin identify users that reported at least one behavior that transgress the norm.
Warning in sorted_count(x): Variable contains value(s) of "" that have been
converted to "empty".| Type | Variable | Missing | Complete Rate | Ordered | No. Unique | Top Counts | Mean (logical) | Counts (logical) | Mean | SD | Min | Perc. 25 | Median | Perc. 75 | Max | Histogram |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| factor | compromiso_biopsicosocial | 0 | 1.0000000 | FALSE | 3 | 2-M: 3593, 3-S: 1976, 1-L: 810 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | dg_cie_10 | 0 | 1.0000000 | FALSE | 2 | 1: 3884, 0: 2495 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | dg_trs_fisico | 0 | 1.0000000 | FALSE | 3 | 1: 3648, 0: 2065, 2: 666 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | abandono_temp | 0 | 1.0000000 | FALSE | 2 | 0: 5163, 1: 1216 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | origen_ingreso_mod | 0 | 1.0000000 | FALSE | 5 | Con: 3441, Sec: 1577, Der: 557, Sec: 516 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | motivodeegreso_mod_imp | 0 | 1.0000000 | FALSE | 7 | Aba: 2047, Alt: 1567, Aba: 1216, Der: 645 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | dg_trs_cons_sus_or | 0 | 1.0000000 | FALSE | 2 | Dep: 4454, Con: 1925 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | diagnostico_trs_fisico | 0 | 1.0000000 | FALSE | 25 | En : 3648, Sin: 2065, Otr: 168, Tra: 115 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | otros_probl_at_sm_or | 1,437 | 0.7747296 | FALSE | 14 | Sin: 2564, Vio: 1587, Otr: 625, Abu: 96 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | sus_principal_mod | 0 | 1.0000000 | FALSE | 5 | Alc: 2468, Pas: 2204, Coc: 1305, Mar: 323 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | otras_sus1_mod | 2,102 | 0.6704813 | FALSE | 5 | Alc: 1752, Mar: 1266, Coc: 732, Pas: 350 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | otras_sus2_mod | 4,109 | 0.3558551 | FALSE | 5 | Mar: 861, Alc: 657, Coc: 434, Pas: 164 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | otras_sus3_mod | 4,109 | 0.3558551 | FALSE | 5 | Mar: 861, Alc: 657, Coc: 434, Pas: 164 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| factor | alta_adm | 0 | 1.0000000 | FALSE | 2 | 0: 5828, 1: 551 | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | hurto | 0 | 1.0000000 | NA | NA | NA | 0.1067565 | FAL: 5698, TRU: 681 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | robo | 0 | 1.0000000 | NA | NA | NA | 0.0608246 | FAL: 5991, TRU: 388 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | venta_drogas | 0 | 1.0000000 | NA | NA | NA | 0.0443643 | FAL: 6096, TRU: 283 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | rina | 0 | 1.0000000 | NA | NA | NA | 0.1250980 | FAL: 5581, TRU: 798 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | otro | 0 | 1.0000000 | NA | NA | NA | 0.0335476 | FAL: 6165, TRU: 214 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | lugar_vivir | 0 | 1.0000000 | NA | NA | NA | 0.9211475 | TRU: 5876, FAL: 503 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | vivienda | 0 | 1.0000000 | NA | NA | NA | 0.9230287 | TRU: 5888, FAL: 491 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | vif | 0 | 1.0000000 | NA | NA | NA | 0.2022261 | FAL: 5089, TRU: 1290 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | transg_norm_bin | 0 | 1.0000000 | NA | NA | NA | 0.3563254 | FAL: 4106, TRU: 2273 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | menor_vivienda | 0 | 1.0000000 | NA | NA | NA | 0.0987616 | FAL: 5749, TRU: 630 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | menor_salud | 0 | 1.0000000 | NA | NA | NA | 0.4400376 | FAL: 3572, TRU: 2807 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | menor_trabajo_edu | 0 | 1.0000000 | NA | NA | NA | 0.4782881 | FAL: 3328, TRU: 3051 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | mayor_total | 0 | 1.0000000 | NA | NA | NA | 0.5063490 | TRU: 3230, FAL: 3149 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | mayor_dosis | 0 | 1.0000000 | NA | NA | NA | 0.5099545 | TRU: 3253, FAL: 3126 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | polycons | 0 | 1.0000000 | NA | NA | NA | 0.8935570 | TRU: 5700, FAL: 679 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | mayor_comp_biopsicsoc | 0 | 1.0000000 | NA | NA | NA | 0.3097664 | FAL: 4403, TRU: 1976 | NA | NA | NA | NA | NA | NA | NA | NA |
| logical | menor_comp_biopsicsoc | 0 | 1.0000000 | NA | NA | NA | 0.1269792 | FAL: 5569, TRU: 810 | NA | NA | NA | NA | NA | NA | NA | NA |
| numeric | total_oh | 0 | 1.0000000 | NA | NA | NA | NA | NA | 7.6126352 | 8.7881215 | 0.0000000 | 0.0000000 | 4.0000000 | 12.0000000 | 28.0000000 | <U+2587><U+2582><U+2582><U+2581><U+2582> |
| numeric | dosis_oh | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.8424518 | 2.1179793 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 39.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_thc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 4.6778492 | 8.9716897 | 0.0000000 | 0.0000000 | 0.0000000 | 4.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | dosis_thc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.8424518 | 2.1179793 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 39.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_pbc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 5.0953127 | 9.1610794 | 0.0000000 | 0.0000000 | 0.0000000 | 6.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | dosis_pbc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 1.1705596 | 2.9942045 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 40.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_coc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 2.6963474 | 6.1555415 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | dosis_coc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.2688509 | 1.7629119 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 40.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_bzd | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.8076501 | 4.0691149 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | dosis_bzd | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.4514814 | 2.8107924 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 40.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_otra | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.9904374 | 4.9582202 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | dosis_otra | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.4514814 | 2.8107924 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 40.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_vif | 0 | 1.0000000 | NA | NA | NA | NA | NA | 1.3500549 | 4.5085774 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_transgresion | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.3705910 | 0.7909335 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 5.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | salud_psicologica | 0 | 1.0000000 | NA | NA | NA | NA | NA | 9.2128860 | 4.6302091 | 0.0000000 | 5.0000000 | 10.0000000 | 12.0000000 | 20.0000000 | <U+2583><U+2586><U+2587><U+2585><U+2581> |
| numeric | total_trabajo | 0 | 1.0000000 | NA | NA | NA | NA | NA | 10.9487380 | 10.5422163 | 0.0000000 | 0.0000000 | 10.0000000 | 20.0000000 | 28.0000000 | <U+2587><U+2581><U+2582><U+2583><U+2583> |
| numeric | total_educacion | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.3099232 | 2.2582653 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 28.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | salud_fisica | 0 | 1.0000000 | NA | NA | NA | NA | NA | 11.1327794 | 4.9691828 | 0.0000000 | 8.0000000 | 10.0000000 | 15.0000000 | 20.0000000 | <U+2582><U+2585><U+2587><U+2586><U+2583> |
| numeric | calidad_vida | 0 | 1.0000000 | NA | NA | NA | NA | NA | 10.4111930 | 5.1114644 | 0.0000000 | 7.0000000 | 10.0000000 | 15.0000000 | 20.0000000 | <U+2583><U+2585><U+2587><U+2585><U+2583> |
| numeric | sin_otros_prob_sm | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.4021006 | 0.4903605 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2586> |
| numeric | dg_cons_dep | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.6982286 | 0.4590626 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | 1.0000000 | <U+2583><U+2581><U+2581><U+2581><U+2587> |
| numeric | dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 1.0039191 | 0.8970966 | 0.0000000 | 0.0000000 | 1.0000000 | 2.0000000 | 2.0000000 | <U+2587><U+2581><U+2583><U+2581><U+2587> |
| numeric | x2_dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0852798 | 0.3780609 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 2.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | x3_dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0227308 | 0.1922389 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 2.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | x4_dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0006271 | 0.0354107 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 2.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | x5_dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | <U+2581><U+2581><U+2587><U+2581><U+2581> |
| numeric | x6_dg_trs_psiq_cie_10_or | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | <U+2581><U+2581><U+2587><U+2581><U+2581> |
| numeric | transg_norm | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.5728171 | 0.9533594 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 6.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | salud_mean | 0 | 1.0000000 | NA | NA | NA | NA | NA | 10.2522861 | 4.1175032 | 0.0000000 | 7.3333333 | 10.0000000 | 13.3333333 | 20.0000000 | <U+2582><U+2585><U+2587><U+2586><U+2582> |
| numeric | dosis_mean | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.6712128 | 1.3873650 | 0.0000000 | 0.0000000 | 0.1666667 | 0.6666667 | 19.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_mean | 0 | 1.0000000 | NA | NA | NA | NA | NA | 3.6467053 | 3.3507752 | 0.0000000 | 0.8333333 | 2.8333333 | 5.3333333 | 23.3333333 | <U+2587><U+2583><U+2581><U+2581><U+2581> |
| numeric | vivienda_mean | 0 | 1.0000000 | NA | NA | NA | NA | NA | 4.6104405 | 1.2392178 | 0.0000000 | 5.0000000 | 5.0000000 | 5.0000000 | 5.0000000 | <U+2581><U+2581><U+2581><U+2581><U+2587> |
| numeric | trabajo_edu_mean | 0 | 1.0000000 | NA | NA | NA | NA | NA | 5.6293306 | 5.3284457 | 0.0000000 | 0.0000000 | 5.0000000 | 10.0000000 | 24.0000000 | <U+2587><U+2582><U+2586><U+2581><U+2581> |
| numeric | total_cie_10 | 0 | 1.0000000 | NA | NA | NA | NA | NA | 1.1125568 | 1.0945562 | 0.0000000 | 0.0000000 | 1.0000000 | 2.0000000 | 8.0000000 | <U+2587><U+2585><U+2581><U+2581><U+2581> |
| numeric | salud_mean_z | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0013466 | 0.9996963 | -2.4878250 | -0.7073514 | -0.0599064 | 0.7493998 | 2.3680123 | <U+2582><U+2585><U+2587><U+2586><U+2582> |
| numeric | dosis_mean_z | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0000000 | 1.0000000 | -0.4838041 | -0.4838041 | -0.3636723 | -0.0032768 | 13.2112223 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | total_mean_z | 0 | 1.0000000 | NA | NA | NA | NA | NA | -0.0007103 | 0.9992102 | -1.0881676 | -0.8396653 | -0.2432600 | 0.5022467 | 5.8698950 | <U+2587><U+2583><U+2581><U+2581><U+2581> |
| numeric | vivienda_mean_z | 0 | 1.0000000 | NA | NA | NA | NA | NA | -0.0002256 | 1.0005570 | -3.7227419 | 0.3143087 | 0.3143087 | 0.3143087 | 0.3143087 | <U+2581><U+2581><U+2581><U+2581><U+2587> |
| numeric | total_vif_z | 0 | 1.0000000 | NA | NA | NA | NA | NA | -0.0009058 | 0.9979097 | -0.2997213 | -0.2997213 | -0.2997213 | -0.2997213 | 5.8976821 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | oh | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.7239379 | 0.4470829 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | 1.0000000 | <U+2583><U+2581><U+2581><U+2581><U+2587> |
| numeric | thc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.3411193 | 0.4741225 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2585> |
| numeric | pbc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.3461358 | 0.4757744 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2585> |
| numeric | coc | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.3049067 | 0.4604040 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2583> |
| numeric | bzd | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0647437 | 0.2460923 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
| numeric | otra | 0 | 1.0000000 | NA | NA | NA | NA | NA | 0.0465590 | 0.2107089 | 0.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | <U+2587><U+2581><U+2581><U+2581><U+2581> |
#knitr::include_graphics("G:/Mi unidad/Alvacast/SISTRAT 2019 (github)/SUD_CL/Figures/Figure_Duplicates.svg")
#foreign::write.dta(CONS_C1_df_dup_JUL_2020_match_top_sel, "G:/Mi unidad/Alvacast/SISTRAT 2019 (github)/c1_top.dta")
#CONS_C1_df_dup_JUL_2020 CONS_C1_df_dup_JUL_2020_match_top CONS_TOP_df_dup_ENE_2020_prev9_match CONS_TOP_df_dup_ENE_2020_prev9
#sólo match
comb_datasets_a_n<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(hash_key_TOP))%>%
dplyr::select(-hash_key_TOP)%>%
nrow()
comb_datasets_a_users<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(hash_key_TOP))%>%
dplyr::select(-hash_key_TOP)%>%
dplyr::distinct(hash_key)%>% nrow()
#sólo los que tienen compromiso biopsicosocial
comb_datasets_b_n<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(compromiso_biopsicosocial), !is.na(hash_key_TOP))%>%
dplyr::select(-hash_key_TOP)%>%
nrow()
comb_datasets_b_users<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(compromiso_biopsicosocial), !is.na(hash_key_TOP))%>%
dplyr::select(-hash_key_TOP)%>%
dplyr::distinct(hash_key)%>% nrow()
#sólo los que tienen etapa de tratamiento de inicio de tratamiento
comb_datasets_c_n<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(compromiso_biopsicosocial), !is.na(hash_key_TOP), etapa_del_tratamiento=="Inicio Tratamiento")%>%
dplyr::select(-hash_key_TOP)%>%
nrow()
comb_datasets_c_users<-CONS_C1_df_dup_JUL_2020%>%
dplyr::mutate(concat=paste0(hash_key,"_",as.character(fech_ing)))%>%
dplyr::left_join(CONS_TOP_df_dup_ENE_2020_prev9_match,by="concat", suffix=c("","_TOP"))%>%
dplyr::filter(!is.na(compromiso_biopsicosocial), !is.na(hash_key_TOP), etapa_del_tratamiento=="Inicio Tratamiento")%>%
dplyr::select(-hash_key_TOP)%>%
dplyr::distinct(hash_key)%>% nrow()
#https://stackoverflow.com/questions/46750364/diagrammer-and-graphviz
#https://mikeyharper.uk/flowcharts-in-r-using-diagrammer/
#http://blog.nguyenvq.com/blog/2012/05/29/better-decision-tree-graphics-for-rpart-via-party-and-partykit/
#http://blog.nguyenvq.com/blog/2014/01/17/skeleton-to-create-fast-automatic-tree-diagrams-using-r-and-graphviz/
#https://cran.r-project.org/web/packages/DiagrammeR/vignettes/graphviz-mermaid.html
#https://stackoverflow.com/questions/39133058/how-to-use-graphviz-graphs-in-diagrammer-for-r
#https://subscription.packtpub.com/book/big_data_and_business_intelligence/9781789802566/1/ch01lvl1sec21/creating-diagrams-via-the-diagrammer-package
#https://justlegal.be/2019/05/using-flowcharts-to-display-legal-procedures/
#
#
library(DiagrammeR) #⋉
grViz("digraph flowchart {
# node definitions with substituted label text
node [fontname = Helvetica, shape = rectangle,fontsize = 10]
tab1 [label = '@@1']
tab2 [label = '@@2']
tab3 [label = '@@3']
tab4 [label = '@@4']
tab5 [label = '@@5']
tab6 [label = '@@6']
tab7 [label = '@@7']
blank [label = '', width = 0.001, height = 0.001]
# edge definitions with the node IDs
tab1 -> tab2;
blank -> tab3[label='',dir = none, shape=none, color = 'white',fontcolor = black, width=0, height=0];
{rank=same; tab2 -> tab3 [label='Left join',fontsize = 11, dir=none]}; #⋉
tab3 -> tab4 [label=' Result of the left join of the dataset',fontsize = 9];
tab4 -> tab5 [label=' Only rows with available data on Biopsychosocial Compromise',fontsize = 9];
tab5 -> tab6 [label=' Only strict matches with applications at the stage of admission',fontsize = 9];
tab6 -> tab7 [label=' Only rows with available data on TOP scores and Diagnostic of CIE-10',fontsize = 9];
}
Only rows with available data on TOP scores
[1]: paste0('TOP Dataset \\n(n = ', formatC(nrow(CONS_TOP_df_dup_ENE_2020_prev9), format='f', big.mark=',', digits=0), ';\\n users: ',formatC(CONS_TOP_df_dup_ENE_2020_prev9%>% dplyr::distinct(hash_key)%>% nrow(), format='f', big.mark=',', digits=0),')')
[2]: paste0('Only applications w/ only one\\n application in the same date \\n(n = ', formatC(nrow(CONS_TOP_df_dup_ENE_2020_prev9_match), format='f', big.mark=',', digits=0), ';\\n users:',formatC(users_top_only_one_app_match, format='f', big.mark=',', digits=0),')')
[3]: paste0('Only applications w/ only one\\n application in the same date \\n(n = ', formatC(nrow(CONS_C1_df_dup_JUL_2020), format='f', big.mark=',', digits=0), ';\\n users:',formatC(CONS_C1_df_dup_JUL_2020%>% dplyr::distinct(hash_key)%>% nrow(), format='f', big.mark=',', digits=0),')')
[4]: paste0('Dataset \\n(n = ',formatC(comb_datasets_a_n,format='f', big.mark=',', digits=0),'\\nusers =',formatC(comb_datasets_a_users,format='f', big.mark=',', digits=0),')')
[5]: paste0('Dataset \\n(n = ',formatC(comb_datasets_b_n,format='f', big.mark=',', digits=0),'\\nusers =',formatC(comb_datasets_b_users,format='f', big.mark=',', digits=0),')')
[6]: paste0('Dataset \\n(n = ',formatC(comb_datasets_c_n,format='f', big.mark=',', digits=0),'\\nusers =',formatC(comb_datasets_c_users,format='f', big.mark=',', digits=0),')')
[7]: paste0('Final Sample \\n(n = ', formatC(nrow(CONS_C1_df_dup_JUL_2020_match_top_sel), format='f', big.mark=',', digits=0), ';\\n users: ',formatC(CONS_C1_df_dup_JUL_2020_match_top_sel%>% dplyr::distinct(hash_key)%>% nrow(), format='f', big.mark=',', digits=0),')')
")To get an idea of the distribution of the biopsychosocial compromise among the selected sample, we generated a pie chart.
library(plotly)
CONS_C1_df_dup_JUL_2020_match_top_sel %>%
dplyr::group_by(compromiso_biopsicosocial)%>%
dplyr::mutate(compromiso_biopsicosocial=case_when(compromiso_biopsicosocial=="2-Moderado"~"Moderate",
compromiso_biopsicosocial=="1-Leve"~"Mild",
TRUE~"Severe"))%>%
dplyr::summarise(n=n(),percent=round(n/sum(n),2))%>%
dplyr::ungroup()%>%
plot_ly( labels = ~compromiso_biopsicosocial, values = ~n, type = 'pie',hole = 0.6,textposition = 'outside',textinfo = 'label+percent', marker = list(colors = c("yellowgreen","orange","tomato"))) %>%
layout(xaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE),
yaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE))Figure 2. Biopsychosocial Compromise
We generated a correlation plot which wraps different type of variables (continuous, polytomous, dichotomous)1. We selected mostly the original variables from the instrument.
# It first identifies the types of variables, organizes them by type (continuous, polytomous, dichotomous), calls the appropriate correlation function, and then binds the resulting matrices together.D
mix_cor<- psych::mixedCor(data=CONS_C1_df_dup_JUL_2020_match_top_sel%>%dplyr::mutate(compromiso_biopsicosocial=as.numeric(compromiso_biopsicosocial)),c=2:20, #total_oh,dosis_oh total_thc dosis_thc total_pbc dosis_pbc total_coc dosis_coc total_bzd dosis_bzd total_otra dosis_otra total_vif total_transgresion salud_psicologica total_trabajo total_educacion salud_fisica calidad_vida
p=1,#compromiso_biopsicosocial,
d=21:27,#, hurto robo venta_drogas rina otro lugar_vivir vivienda
smooth=TRUE,
correct=.5,
global=TRUE,
ncat=8,
use="pairwise",
method="kendall",
weight=NULL)# Getting the correlation matrix of the dataset.
comp_biopsicosoc_mat <- mix_cor$rho
# Explore the correlation structure of the dataset.
library(ggcorrplot)
ggcorrplot(comp_biopsicosoc_mat,
ggtheme = ggplot2::theme_gray,
colors = c("#E46726", "white", "#6D9EC1"),
tl.cex=8)
Figure 3.Matrix Of Correlations Between Original Variables
#subscript out of bounds
#mixedCor(data=CONS_C1_df_dup_JUL_2020_match_top_sel,c=c("total_oh","dosis_oh", "total_thc", "dosis_thc", "total_pbc", "dosis_pbc", "total_coc", "dosis_coc", "total_bzd", "dosis_bzd", "total_otra", "dosis_otra", "total_vif", "total_transgresion", "salud_psicologica", "total_trabajo", "total_educacion", "salud_fisica", "calidad_vida"), p="compromiso_biopsicosocial",d=c("hurto", "robo", "venta_drogas", "rina", "otro", "lugar_vivir", "vivienda"), smooth=TRUE,correct=.5,global=TRUE,ncat=8,use="pairwise",method="pearson",weight=NULL)
#subscript out of bounds
#set.seed(123)
#par(mar=c(1,1,1,1))
#comp_biopsicosoc_mat_ci<-mycor.ci(x=CONS_C1_df_dup_JUL_2020_match_top_sel%>%dplyr::mutate(compromiso_biopsicosocial=as.numeric(compromiso_biopsicoso#cial)),
# method="spearman",
# poly=TRUE,
# cvars=2:20,
# pvars=1,
# dvars=21:27,
# n.iter=1000,
# plot = F)
#
#cor_pmat(): para ver la sig estadística de las correlacionesAlso we included an heterogeneus matrix of correlations, consisting of polychoric, biserial and pearson correlations.2 We mostly selected transformed variables in the following matrix.
require(polycor)
#Computes a heterogenous correlation matrix, consisting of Pearson product-moment correlations between numeric variables, polyserial correlations between numeric and ordinal variables, and polychoric correlations between ordinal variables.
hetcor_mat<-hetcor(CONS_C1_df_dup_JUL_2020_match_top_sel[c("compromiso_biopsicosocial","lugar_vivir","vivienda","transg_norm","salud_mean","dosis_mean","total_mean","trabajo_edu_mean","vivienda_mean","polycons", "menor_salud", "menor_trabajo_edu", "mayor_dosis", "mayor_total","transg_norm_bin","dg_cons_dep", "sin_otros_prob_sm","abandono_temp","alta_adm","dg_cie_10")], ML = T, std.err = T, use="complete.obs", bins=4, pd=TRUE)
#CONS_C1_df_dup_JUL_2020_match_top_sel%>% select(contains("dosis"))%>% hist()
ggcorrplot(hetcor_mat$correlations,
ggtheme = ggplot2::theme_gray,
colors = c("#E46726", "white", "#6D9EC1"), tl.cex=8)
Figure 4.Matrix Of Correlations Between Transformed Variables
#pcor(c("a", "b", "x", "y", "z"), var(mydata))
# partial corr between a and b controlling for x, y, zVariables of TOP depending on Biopsychosocial Compromise
library(reshape2) # for melt() function
Attaching package: 'reshape2'The following objects are masked from 'package:data.table':
dcast, meltThe following object is masked from 'package:tidyr':
smithslibrary(ggplot2)
# First we need to restructure the data into long format:
ocdMelt <- melt(CONS_C1_df_dup_JUL_2020_match_top_sel[c('compromiso_biopsicosocial','hurto',"robo","venta_drogas","rina","otro","lugar_vivir","vivienda","polycons","vif","transg_norm_bin")], id=c('compromiso_biopsicosocial'))
names(ocdMelt) <- c('Group', 'Outcome_Measure', 'Frequency')
ocdMelt <-ocdMelt%>% dplyr::mutate(Frequency=factor(Frequency))%>% dplyr::group_by(Outcome_Measure,Group,Frequency)%>% dplyr::arrange(Outcome_Measure, Group,Frequency)%>% dplyr::add_tally()%>% distinct(.,.keep_all=T)%>% dplyr::ungroup()%>% dplyr::group_by(Group, Outcome_Measure)%>% dplyr::mutate(perc=n/sum(n))%>% dplyr::ungroup()
# plot
p5<-ggplot(ocdMelt, aes(Outcome_Measure,perc))+
geom_bar(aes(fill= Frequency),stat="identity")+
facet_grid(Group~.)+
scale_y_continuous(labels = scales::percent_format(accuracy = 5L),limits=c(0,1))+
scale_fill_manual(values=c("grey60","steelblue"),labels=c("Absence", "Presence")) +
sjPlot::theme_sjplot2() +
ylab("Biopsychosocial Compromise")+
theme(axis.text.x = element_text(size=8,vjust = .5,angle=25), axis.text.y = element_text(size=10),
axis.title.y = element_text(size = 14)) +
theme(
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.x = element_blank(),
panel.background = element_blank(),
line = element_blank())
p5
Figure 5. Bar Plot of Binary Answers By Biopsychosocial Compromise
We generated z scores for those composite scores of multiple variables, in order to make them comparable. In the following figure we may observe the distribution of continuous variables among users with different levels of biopsychosocial compromise.
library(reshape2) # for melt() function
library(ggplot2)
# First we need to restructure the data into long format:
ocdMelt <- melt(CONS_C1_df_dup_JUL_2020_match_top_sel[c('compromiso_biopsicosocial',"salud_mean_z","dosis_mean_z", "total_mean_z","total_vif_z")], id=c('compromiso_biopsicosocial'))Warning: attributes are not identical across measure variables; they will be
droppednames(ocdMelt) <- c('Group', 'Outcome_Measure', 'Frequency')
# plot
ocdBoxplot <- ggplot(ocdMelt, aes(Group, Frequency, color = Outcome_Measure)) +
geom_jitter(alpha=0.05) +
geom_boxplot() +
labs(x='Biopsychosocial Compromise', y='Scores', color='Variable') + scale_y_continuous(breaks=seq(0,50, by=5))+
sjPlot::theme_sjplot2() +
scale_color_manual(values=c("darkslategray4","rosybrown","dimgray","coral3","grey60"),na.value="black") + #"Set2"
theme(
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.x = element_blank(),
panel.background = element_blank(),
line = element_blank())
ocdBoxplot
Figure 6. Box Plot of Continuous Answers By Biopsychosocial Compromise
As seen above, users categorized with a severe Biopsychosocial Compromise showed slightly greater patterns of consumption and lower average health in comparison to users with a mild Biopsychosocial Compromise. However, and at least in an univariate level, they seem to not differ much between the different levels.
We checked the multivariate normality of these variables, in order to choose the technique to perform the analyses that may let us infer whether that users in different levels of Biopsychosocial Compromise provene from the same population.
| Test | Statistic | P Value | Normality |
|---|---|---|---|
| Mardia Skewness | 49,049.954 | 0 | NO |
| Mardia Kurtosis | 350.767 | 0 | NO |
| MVN | NA | NA | NO |
| Note. Using z-scores distributions of Average Health, Average Frequency of Use in the last Month and Average Dose in the las Month |
As seen above, variables did not follow a normal distribution.
Non-parametric Multivariate Test
Considering that our variables did not follow a normal distribution, and the test is composed by a mixture of different variable types, we performed a nonparametric test for multivariate samples. The variability of responses on each of the variables of top was compared between the groups using a non-parametric multivariate ANOVA (MANOVA (Burchett et al., 2017)) with biopsychosocial compromise as between-subject factor (discarding inter-observer variability). The procedure generates a randomization test of 5,000 resamples to compute F-statistic (pseudo-F). Additionally, these models have the advantage that allow for differences in between-group variation, is not sensitive to multicollinearity, allows for the inclusion of more variables, and tolerate sparse data.
#mvnormtest::mshapiro.test()
dependent.vars <- cbind(CONS_C1_df_dup_JUL_2020_match_top_sel$salud_mean_z,CONS_C1_df_dup_JUL_2020_match_top_sel$dosis_mean_z, CONS_C1_df_dup_JUL_2020_match_top_sel$total_mean_z)
root.manova <- manova(dependent.vars ~ CONS_C1_df_dup_JUL_2020_match_top_sel$compromiso_biopsicosocial, intercept=F)
summary(root.manova,test="Pillai")
#test='Pillai'). The manova() function provides four multivariate tests by setting the test argument to either Pillai, #Wilks, Hotelling-Lawley, or Roy.
#La prueba de Wilk es la más comúnmente utilizada debido a que fue la primera prueba en ser derivada y a que posee una aproximación F bastante conocida.
#Lawley-Hotelling es también conocida como el estadístico T2 generalizado de Hotelling.
#La prueba de Pillai es la mejor prueba a utilizar en la mayoría de las situaciones. La prueba de Pillai proporciona resultados similares a los de las pruebas de Wilks y de Lawley-Hotelling.
#La prueba de la raíz más grande de Roy es la mejor cuando los vectores medios son colineales. La prueba de Roy no tiene una aproximación F satisfactoria.
summary.aov(root.manova)CONS_C1_df_dup_JUL_2020_match_top_sel2<-
CONS_C1_df_dup_JUL_2020_match_top_sel%>%
dplyr::mutate_at(vars(hurto, robo, venta_drogas, rina, otro, lugar_vivir, vivienda,vif), ~as.numeric(as.factor(as.numeric(.))))
#https://stackoverflow.com/questions/59166335/need-help-using-npmv-package-nonparametric-testing-in-r
#https://cran.r-project.org/web/packages/npmv/npmv.pdf
#R.Ellis, A., W.Burchett, W., W.Harrar, S., & Bathke, A. (2017). Nonparametric Inference for Multivariate Data: the Package npmv. JOURNAL OF STATISTICAL SOFTWARE. https://doi.org/10.18637/jss.v076.i04
invisible(
prueba_anormalidad1<-npmv::nonpartest(dosis_mean_z|total_mean_z|salud_mean_z|hurto|robo|venta_drogas|rina|otro|lugar_vivir|vivienda|total_vif_z~compromiso_biopsicosocial, CONS_C1_df_dup_JUL_2020_match_top_sel2,plots=F,permreps=5000,releffects=TRUE)
)
invisible(
prueba_anormalidad2<-npmv::nonpartest(dosis_mean_z|total_mean_z|salud_mean_z|hurto|robo|venta_drogas|rina|otro|lugar_vivir|vivienda|total_vif_z~compromiso_biopsicosocial+origen_ingreso_mod, CONS_C1_df_dup_JUL_2020_match_top_sel2,plots=F,permreps=5000,releffects=TRUE)
)
invisible(
prueba_anormalidad3<-npmv::nonpartest(dosis_mean_z|total_mean_z|salud_mean_z|hurto|robo|venta_drogas|rina|otro|lugar_vivir|vivienda|total_vif_z~compromiso_biopsicosocial+origen_ingreso_mod+dg_trs_fisico, CONS_C1_df_dup_JUL_2020_match_top_sel2,plots=F,permreps=5000,releffects=TRUE)
)
invisible(
prueba_anormalidad4<-npmv::nonpartest(dosis_mean_z|total_mean_z|salud_mean_z|hurto|robo|venta_drogas|rina|otro|lugar_vivir|vivienda|total_vif_z~ compromiso_biopsicosocial+origen_ingreso_mod+dg_trs_fisico+dg_cie_10, CONS_C1_df_dup_JUL_2020_match_top_sel2,plots=F,permreps=5000,releffects=TRUE)
)
#
prueba_anormalidad_res1<-data.table::data.table(prueba_anormalidad1$results, keep.rownames=T)
prueba_anormalidad_res2<-data.table::data.table(prueba_anormalidad2$results, keep.rownames=T)
prueba_anormalidad_res3<-data.table::data.table(prueba_anormalidad3$results, keep.rownames=T)
prueba_anormalidad_res4<-data.table::data.table(prueba_anormalidad4$results, keep.rownames=T)
#INOCRPORË LAS VAR DE CONTROL. SON TODAS DE TIPO FACTOR
sum_df_npmv<-data.frame(cbind(Model=
paste0("npmv",c("Raw","Raw","Raw","Raw","2","2","2","2","3","3","3","3","4","4","4","4")),
rbind(prueba_anormalidad_res1,
prueba_anormalidad_res2,
prueba_anormalidad_res3,
prueba_anormalidad_res4)))%>%
dplyr::rename("Parameter"="rn")
sum_df_npmv%>%
dplyr::mutate(Test.Statistic=round(as.numeric(as.character(Test.Statistic)),3))%>%
knitr::kable(.,format = "html", format.args = list(decimal.mark = ".", big.mark = ","),
caption="Table 3. Analysis of Variance",
align =rep('c', 101),
col.names=c("Model", "Parameter", "Statistic", "DF1", "DF2","p-value", "Permutation Test p-value"))%>%
kableExtra::kable_styling(bootstrap_options = c("striped", "hover"),font_size = 8) %>%
kableExtra::add_footnote( c("Note. Model with average dose (z score), average frequency of use (z score), theft, burglary, drug selling, fights, other, place to live, housing condition and domestic violence. Raw= No control variables; Controlling for Motive of Admission(1); Controlling for Motive of Admission & Physical Condition(2); Controlling for Motive of Admission, Physical Condition & Psychiatric Condition(3)"), notation = "none") %>%
kableExtra::scroll_box(width = "100%", height = "375px")| Model | Parameter | Statistic | DF1 | DF2 | p-value | Permutation Test p-value |
|---|---|---|---|---|---|---|
| npmvRaw | ANOVA type test p-value | 165.153 | 10.210 | 22,736.373 | 0 | 0 |
| npmvRaw | McKeon approx. for the Lawley Hotelling Test | 48.867 | 22.000 | 10,909.062 | 0 | 0 |
| npmvRaw | Muller approx. for the Bartlett-Nanda-Pillai Test | 45.850 | 22.010 | 12,737.991 | 0 | 0 |
| npmvRaw | Wilks Lambda | 47.360 | 22.000 | 12,732.000 | 0 | 0 |
| npmv2 | ANOVA type test p-value | 18.621 | 23.928 | 18,350.079 | 0 | 0 |
| npmv2 | McKeon approx. for the Lawley Hotelling Test | 11.546 | 48.000 | 18,701.236 | 0 | 0 |
| npmv2 | Muller approx. for the Bartlett-Nanda-Pillai Test | 11.458 | 48.041 | 25,481.970 | 0 | 0 |
| npmv2 | Wilks Lambda | 11.507 | 48.000 | 24,512.980 | 0 | 0 |
| npmv3 | ANOVA type test p-value | 16.327 | 13.184 | 26,222.112 | 0 | 0 |
| npmv3 | McKeon approx. for the Lawley Hotelling Test | 7.570 | 26.000 | 11,132.688 | 0 | 0 |
| npmv3 | Muller approx. for the Bartlett-Nanda-Pillai Test | 7.508 | 26.012 | 12,733.989 | 0 | 0 |
| npmv3 | Wilks Lambda | 7.539 | 26.000 | 12,728.000 | 0 | 0 |
| npmv4 | ANOVA type test p-value | 80.904 | 7.610 | 46,227.035 | 0 | 0 |
| npmv4 | McKeon approx. for the Lawley Hotelling Test | 34.028 | 14.000 | 6,364.000 | 0 | 0 |
| npmv4 | Muller approx. for the Bartlett-Nanda-Pillai Test | 34.023 | 14.002 | 6,363.998 | 0 | 0 |
| npmv4 | Wilks Lambda | 34.028 | 14.000 | 6,364.000 | 0 | 0 |
| Note. Model with average dose (z score), average frequency of use (z score), theft, burglary, drug selling, fights, other, place to live, housing condition and domestic violence. Raw= No control variables; Controlling for Motive of Admission(1); Controlling for Motive of Admission & Physical Condition(2); Controlling for Motive of Admission, Physical Condition & Psychiatric Condition(3) |
##NO OCUPAR: MUY COMPLICADO
#The function ssnonpartest performs an all-subset algorithm to determine which variables cause significant effects, and between which factor levels. See the examples below for some basic uses and look in the help pages for each function for a much more detailed look.
#invisible(
#npmv::ssnonpartest(dosis_mean|total_mean|trabajo_edu_mean~compromiso_biopsicosocial,CONS_C1_df_dup_JUL_2020_match_top_sel,test=c(1,0,0,0),alpha=.05,factors.and.variables=TRUE)
#)#https://www.rdocumentation.org/packages/vegan/versions/2.4-2/topics/adonis
#t looks like the jaccard distance is really only useful for binary data (presence/absence) while The bray-curtis matrix has been found to be robust for many abundance type data sets, especially those with many paired zeros like stomach content data can have.
vegan_1<-
vegan::adonis(CONS_C1_df_dup_JUL_2020_match_top_sel2[,c("dosis_mean_z", "total_mean_z", "salud_mean_z", "hurto", "robo", "venta_drogas", "rina", "otro", "lugar_vivir", "vivienda", "total_vif_z"),drop = FALSE] ~ compromiso_biopsicosocial, data=CONS_C1_df_dup_JUL_2020_match_top_sel,permutations=1000, parallel = 4)Warning in vegdist(lhs, method = method, ...): results may be meaningless
because data have negative entries in method "bray"vegan_2<-
vegan::adonis(CONS_C1_df_dup_JUL_2020_match_top_sel2[,c("dosis_mean_z", "total_mean_z", "salud_mean_z", "hurto", "robo", "venta_drogas", "rina", "otro", "lugar_vivir", "vivienda", "total_vif_z"),drop = FALSE] ~ compromiso_biopsicosocial+origen_ingreso_mod, data=CONS_C1_df_dup_JUL_2020_match_top_sel,permutations=1000, parallel = 4)Warning in vegdist(lhs, method = method, ...): results may be meaningless
because data have negative entries in method "bray"vegan_3<-
vegan::adonis(CONS_C1_df_dup_JUL_2020_match_top_sel2[,c("dosis_mean_z", "total_mean_z", "salud_mean_z", "hurto", "robo", "venta_drogas", "rina", "otro", "lugar_vivir", "vivienda", "total_vif_z"),drop = FALSE] ~ compromiso_biopsicosocial+origen_ingreso_mod+dg_trs_fisico, data=CONS_C1_df_dup_JUL_2020_match_top_sel, permutations=1000, parallel = 4)Warning in vegdist(lhs, method = method, ...): results may be meaningless
because data have negative entries in method "bray"vegan_4<-
vegan::adonis(CONS_C1_df_dup_JUL_2020_match_top_sel2[,c("dosis_mean_z", "total_mean_z", "salud_mean_z", "hurto", "robo", "venta_drogas", "rina", "otro", "lugar_vivir", "vivienda", "total_vif_z"), drop = FALSE] ~ compromiso_biopsicosocial+origen_ingreso_mod+dg_trs_fisico+dg_cie_10, data=CONS_C1_df_dup_JUL_2020_match_top_sel, permutations=1000, parallel = 4)Warning in vegdist(lhs, method = method, ...): results may be meaningless
because data have negative entries in method "bray"sum_df_vegan<-data.frame(cbind(Model=paste0("vegan",c("Raw","Raw","Raw",
"2","2","2","2",
"3","3","3","3","3",
"4","4","4","4","4","4")),
rbind(data.table::data.table(vegan_1$ aov.tab, keep.rownames=T),
data.table::data.table(vegan_2$ aov.tab, keep.rownames=T),
data.table::data.table(vegan_3$ aov.tab, keep.rownames=T),
data.table::data.table(vegan_4$ aov.tab, keep.rownames=T))))%>%
dplyr::rename("Parameter"="rn")
sum_df_vegan%>%
dplyr::mutate(R2=round(as.numeric(as.character(R2)),3))%>%
knitr::kable(.,format = "html", format.args = list(decimal.mark = ".", big.mark = ","),
caption="Table 4. Analysis of Variance using bray-curtis matrix",
align =rep('c', 101),
col.names=c("Model","Parameter", "Df","SS", "Mean Squares","Pseudo F","R2", "p-value"))%>%
kableExtra::kable_styling(bootstrap_options = c("striped", "hover"),font_size = 8) %>%
kableExtra::add_footnote( c("Note. Model with average dose (z score), average frequency of use (z score), theft, burglary, drug selling, fights, other, place to live, housing condition and domestic violence. Raw= No control variables; Controlling for Motive of Admission(1); Controlling for Motive of Admission & Physical Condition(2); Controlling for Motive of Admission, Physical Condition & Psychiatric Condition(3)"), notation = "none") %>%
kableExtra::scroll_box(width = "100%", height = "375px")| Model | Parameter | Df | SS | Mean Squares | Pseudo F | R2 | p-value |
|---|---|---|---|---|---|---|---|
| veganRaw | compromiso_biopsicosocial | 2 | 11.4921579 | 5.7460790 | 172.839357 | 0.051 | 0.000999 |
| veganRaw | Residuals | 6,376 | 211.9713947 | 0.0332452 | NA | 0.949 | NA |
| veganRaw | Total | 6,378 | 223.4635527 | NA | NA | 1.000 | NA |
| vegan2 | compromiso_biopsicosocial | 2 | 11.4921579 | 5.7460790 | 174.720466 | 0.051 | 0.000999 |
| vegan2 | origen_ingreso_mod | 4 | 2.4137164 | 0.6034291 | 18.348410 | 0.011 | 0.000999 |
| vegan2 | Residuals | 6,372 | 209.5576783 | 0.0328873 | NA | 0.938 | NA |
| vegan2 | Total | 6,378 | 223.4635527 | NA | NA | 1.000 | NA |
| vegan3 | compromiso_biopsicosocial | 2 | 11.4921579 | 5.7460790 | 175.033627 | 0.051 | 0.000999 |
| vegan3 | origen_ingreso_mod | 4 | 2.4137164 | 0.6034291 | 18.381297 | 0.011 | 0.000999 |
| vegan3 | dg_trs_fisico | 2 | 0.4405869 | 0.2202934 | 6.710447 | 0.002 | 0.000999 |
| vegan3 | Residuals | 6,370 | 209.1170914 | 0.0328284 | NA | 0.936 | NA |
| vegan3 | Total | 6,378 | 223.4635527 | NA | NA | 1.000 | NA |
| vegan4 | compromiso_biopsicosocial | 2 | 11.4921579 | 5.7460790 | 175.331041 | 0.051 | 0.000999 |
| vegan4 | origen_ingreso_mod | 4 | 2.4137164 | 0.6034291 | 18.412530 | 0.011 | 0.000999 |
| vegan4 | dg_trs_fisico | 2 | 0.4405869 | 0.2202934 | 6.721849 | 0.002 | 0.000999 |
| vegan4 | dg_cie_10 | 1 | 0.3874970 | 0.3874970 | 11.823759 | 0.002 | 0.000999 |
| vegan4 | Residuals | 6,369 | 208.7295944 | 0.0327727 | NA | 0.934 | NA |
| vegan4 | Total | 6,378 | 223.4635527 | NA | NA | 1.000 | NA |
| Note. Model with average dose (z score), average frequency of use (z score), theft, burglary, drug selling, fights, other, place to live, housing condition and domestic violence. Raw= No control variables; Controlling for Motive of Admission(1); Controlling for Motive of Admission & Physical Condition(2); Controlling for Motive of Admission, Physical Condition & Psychiatric Condition(3) |
#dg_cons_dep", "sin_otros_prob_sm","abandono_temp",,"dg_cie_10"
#https://stackoverflow.com/questions/16638297/vegdist-error-message-in-vegan-packageRelationship with outcomes
library(glmulti)
#sin_otros_prob_sm+ dg_cons_dep
res <- glmulti(abandono_temp ~ compromiso_biopsicosocial + origen_ingreso_mod + dg_cie_10 + dg_trs_fisico, #tal vez no debería meter esta última variable.
data=CONS_C1_df_dup_JUL_2020_match_top_sel2%>%dplyr::mutate(compromiso_biopsicosocial=case_when(compromiso_biopsicosocial=="2-Moderado"~"Moderate",compromiso_biopsicosocial=="1-Leve"~"Mild",TRUE~"Severe")),
level=2, #The level of interaction between explanatory variables to be considered. Currently only 1 (only main effects) or 2 (main effects plus all pairwise interactions) are supported.
fitfunction=glm, crit="aicc",
family=binomial,
marginality=T,#Whether to use the general marginality rule or not. Default is FALSE. With TRUE, all predictors found in interaction terms are also included as main effects.
confsetsize = 1000 #How many models should be returned in the confidence set of models?
)Initialization...
TASK: Exhaustive screening of candidate set.
Fitting...
After 50 models:
Best model: abandono_temp~1+origen_ingreso_mod+dg_cie_10+dg_trs_fisico
Crit= 5789.70750899756
Mean crit= 5942.2170841595
After 100 models:
Best model: abandono_temp~1+origen_ingreso_mod+dg_cie_10+dg_trs_fisico+compromiso_biopsicosocial
Crit= 5779.5616968649
Mean crit= 5866.2943217028Completed.#t is not iterative (all models are compared), and is fully automatic (no intervention of the user is required). It is primarily intended to be used for exploratory analyses with many candidate explanatory variables (say 10 to 20)
plot(res)##########################################################################################
##################DEBIESE INCORPORAR EL TIPO DE PLAN!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
##########################################################################################
#http://www.metafor-project.org/doku.php/tips:model_selection_with_glmulti_and_mumin
#But let's take a look at the top 10 models:
top <- weightable(res) #Prepares a table with model formulas, IC values and IC relative supports
top <- top[top$aicc <= min(top$aicc) + 5,]
############################################################################################
###########################################################################################
plot.glmulti<-function(x, type="p", ...)
{
if ( type == "p") {
plot(x@crits,xlab="Best models",
ylab=paste("Support (",x@params$crit,")"),pch=19, main="IC profile", ...)
abline(h=x@crits[1]+2,col="red")
return(assign("ic_glmulti",data.table::data.table(data.frame(imp),keep.rownames = T),envir = .GlobalEnv))
} else if (type == "w") {
ww = exp(-(x@crits - x@crits[1])/2)
ww=ww/sum(ww)
plot(ww,xlab="Best models",
ylab=paste("Evidence weight (",x@params$crit,")"),pch=19, main="Profile of model weights", ...)
cucu=function(i) sum(ww[1:i])
wwc=lapply(1:length(ww),cucu)
abline(v=min(which(wwc>=0.95)),col="red")
} else if (type=="r") {
if (length(x@objects)) {
# shows some diagnostics of the fit
dev.new()
par(mfrow=c(2,min(length(x@crits), 5)))
for (k in 1:min(length(x@crits), 5))
plot(x@objects[[k]],which=c(1), main=deparse(x@formulas[[k]]),...)
for (k in 1:min(length(x@crits), 5))
plot(x@objects[[k]],which=c(2),...)
} else warning("Unavailable: use includeobjects=T when calling glmulti.")
} else if (type=="s") {
# plots variable (i.e. terms) importances
ww = exp(-(x@crits - x@crits[1])/2)
ww=ww/sum(ww)
# handle synonymies for interactions
# this translates to unique notations (e.g. x:y and y:x are the same)
clartou=function(x) {
sort(strsplit(x, ":")[[1]])-> pieces
if (length(pieces)>1) paste(pieces[1],":",pieces[2],sep="")
else x
}
# list terms in models
tet = lapply(x@formulas, function(x) sapply(attr(delete.response(terms(x)),"term.labels"), clartou))
# all unique terms
unique(unlist(tet))-> allt
# importances
sapply(allt, function(x) sum(ww[sapply(tet, function(t) x%in%t)]))-> imp
# draw
#par(oma=c(0,13,0,0))
barplot(sort(imp),xlab="",xlim=c(0,1), ylab="",horiz=T,las=2, names.arg=allt[order(imp)],main="Model-averaged importance of terms", ...)
abline(v=0.8, col="red")
return(assign("var_imp_glmulti",data.table::data.table(data.frame(imp),keep.rownames = T),envir = .GlobalEnv))
} else warning("plot: Invalid type argument for plotting glmulti objects.")
}
plot(res, type="s")############################################################################################data.frame(res@crits)%>% dplyr::mutate(rn=row_number())%>%
ggplot(aes(x=rn,y=res.crits))+
geom_point(color="steelblue")+
sjPlot::theme_sjplot2() +
ylab("Support (aicc)")+
theme(axis.text.x = element_text(size=8,vjust = .5,angle=25), axis.text.y = element_text(size=10),
axis.title.y = element_text(size = 14)) +
theme(
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.x = element_blank(),
panel.background = element_blank(),
line = element_blank())+
geom_hline(yintercept=res@crits[1]+2, color="red", linetype=2,size=1)+
xlab("Best models")
Figure 7. Ordered Models and AICC
We selected the model with lower AIC. Of them, the first selected did not contain biopsychosocial compromise (AICc= 5779.53, IC Rel. Support= 0.144). The second contained less interaction terms (did not contain the interaction between cause of admission and biopsychosocial compromise) (AICc= 5780.21, IC Rel. Support= 0.101).
#par(oma=c(0,0,0,24))
ggplot(var_imp_glmulti, aes(x=reorder(rn, -imp),y=imp))+
geom_bar(stat="identity",alpha=.7, fill="steelblue")+
sjPlot::theme_sjplot2() +
ylab("Model-averaged importance")+
theme(axis.text.x = element_text(size=6.5,vjust = .5,angle=25),
axis.text.y = element_text(size=10),
axis.title.y = element_text(size = 14)) +
theme(
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.x = element_blank(),
panel.background = element_blank(),
line = element_blank())+
#geom_hline(yintercept=.5, color="red", linetype=2)+
xlab("Terms")
Figure 8. Model-averaged variable importance of terms
df_coef_glmulti<- data.frame()
for (i in 1:nrow(summary(res@objects[[2]])$coefficients)){
df<-cbind(attr(summary(res@objects[[2]])$coefficients,"dimnames")[[1]][i],
t(data.frame(exp(summary(res@objects[[2]])$coefficients[i,1] +
+ qnorm(c(0.025,.5,0.975)) * summary(res@objects[[2]])$coefficients[i,2]))))
df_coef_glmulti<-rbind(df_coef_glmulti,df)
}
df_coef_glmulti<-data.table::data.table(df_coef_glmulti)
data.table::data.table(summary(res@objects[[2]])$coefficients,keep.rownames = T)%>%
dplyr::left_join(df_coef_glmulti,by=c("rn"="V1"))%>%
dplyr::select(-V3)%>%
dplyr::mutate(V2=round(as.numeric(as.character(V2)),2))%>%
dplyr::mutate(V4=round(as.numeric(as.character(V4)),2))%>%
dplyr::mutate(Estimate=round(as.numeric(as.character(Estimate)),2))%>%
dplyr::mutate(OR=round(exp(as.numeric(as.character(Estimate))),2))%>%
#exp(summary(m)$coefficients["DSH",1] +
#+ qnorm(c(0.025,0.5,0.975)) * summary(m)$coefficients["DSH",2])
#[1] 1.472098 4.197368 11.967884
dplyr::mutate(`Std. Error`=round(as.numeric(as.character(`Std. Error`)),2))%>%
dplyr::mutate(`z value`=round(as.numeric(as.character(`z value`)),2))%>%
dplyr::mutate(`Pr(>|z|)`=round(as.numeric(as.character(`Pr(>|z|)`)),4))%>%
dplyr::rename("Terms"="rn", "OR_lo"="V2","OR_up"="V4")%>%
dplyr::select(Terms,Estimate,OR,OR_lo,OR_up, everything())%>%
knitr::kable(.,format = "html", format.args = list(decimal.mark = ".", big.mark = ","),
caption="Table 6. Logistic Regression Coefficients",
align =rep('c', 101))%>%
kableExtra::kable_styling(bootstrap_options = c("striped", "hover"),font_size = 8)%>%
kableExtra::add_footnote( c(paste("Note. AICc=",round(res@objects[[2]]$aic,3)),
paste("Defined model=",top$model[2])), notation = "none")%>%
kableExtra::scroll_box(width = "100%", height = "375px")| Terms | Estimate | OR | OR_lo | OR_up | Std. Error | z value | Pr(>|z|) |
|---|---|---|---|---|---|---|---|
| (Intercept) | -2.17 | 0.11 | 0.08 | 0.16 | 0.17 | -13.03 | 0.0000 |
| origen_ingreso_modDerivación Asistida | -0.23 | 0.79 | 0.62 | 1.01 | 0.12 | -1.92 | 0.0554 |
| origen_ingreso_modOtros | -0.50 | 0.61 | 0.43 | 0.86 | 0.18 | -2.77 | 0.0057 |
| origen_ingreso_modSector Justicia | -0.34 | 0.71 | 0.55 | 0.92 | 0.13 | -2.61 | 0.0090 |
| origen_ingreso_modSector Salud | -0.36 | 0.70 | 0.60 | 0.82 | 0.08 | -4.37 | 0.0000 |
| dg_cie_101 | 1.49 | 4.44 | 3.76 | 5.27 | 0.09 | 17.34 | 0.0000 |
| dg_trs_fisico1 | -0.68 | 0.51 | 0.33 | 0.79 | 0.22 | -3.04 | 0.0023 |
| dg_trs_fisico2 | -0.81 | 0.44 | 0.18 | 1.12 | 0.47 | -1.72 | 0.0846 |
| compromiso_biopsicosocialModerate | 0.03 | 1.03 | 0.73 | 1.45 | 0.17 | 0.19 | 0.8520 |
| compromiso_biopsicosocialSevere | 0.08 | 1.08 | 0.75 | 1.59 | 0.19 | 0.44 | 0.6626 |
| dg_trs_fisico1:compromiso_biopsicosocialModerate | 0.29 | 1.34 | 0.84 | 2.16 | 0.24 | 1.22 | 0.2242 |
| dg_trs_fisico2:compromiso_biopsicosocialModerate | 0.24 | 1.27 | 0.47 | 3.37 | 0.50 | 0.47 | 0.6378 |
| dg_trs_fisico1:compromiso_biopsicosocialSevere | 0.57 | 1.77 | 1.07 | 2.93 | 0.26 | 2.23 | 0.0258 |
| dg_trs_fisico2:compromiso_biopsicosocialSevere | 0.13 | 1.14 | 0.42 | 3.08 | 0.51 | 0.26 | 0.7920 |
| Note. AICc= 5780.205 | |||||||
| Defined model= abandono_temp ~ 1 + origen_ingreso_mod + dg_cie_10 + dg_trs_fisico + compromiso_biopsicosocial + dg_trs_fisico:compromiso_biopsicosocial |
#abandono_temp ~ 1 + compromiso_biopsicosocial + origen_ingreso_mod + dg_cie_10 + dg_trs_fisico + dg_trs_fisico:compromiso_biopsicosocial
#print(res@formulas[[2]])The next figures show the marginal effects of the interaction terms among the variables.
library(lsmeans)
refR_1 <- lsmeans(res@objects[[2]], specs = c("compromiso_biopsicosocial","dg_trs_fisico"))
ref_dfR_1 <- as.data.frame(cbind(summary(refR_1)[1:2],exp(summary(refR_1)[,c(3,6,7)])))
g4R_1 <- ggplot(ref_dfR_1, aes(x=compromiso_biopsicosocial, y=lsmean,group=dg_trs_fisico, colour=dg_trs_fisico))+
geom_errorbar(aes(ymin=asymp.LCL, ymax=asymp.UCL), width=.1,position=position_dodge(0.1), size=1) +
geom_line(position=position_dodge(0.1), size=.5)+
geom_point(position=position_dodge(0.1), size=2)+
xlab("Biopsychosocial Compromise")+
ylab("Odds Ratio (OR) of Early Drop-out")+
sjPlot::theme_sjplot2() +
geom_rect_interactive(alpha = 0.1, xmin=.1, xmax=.1, ymin=.1,ymax=.1) +
# Remove plot elements added by geom_rect_interactive
theme(legend.position="bottom")+
guides(color=guide_legend(ncol=4))+
labs(color="Cause of Discharge")+
scale_colour_brewer(name="Cumulative Percentage", palette = "Set1",labels=c("No Dg.", "In study", "Dg."))+
theme(legend.title = element_blank())
#theme(axis.text = element_blank(), axis.ticks = element_blank())
ggiraph(code = print(g4R_1))Figure 9. Interactive Effects in the Prob. of Early Drop-outv with Diagnose of Physical Diagnosis
library(lsmeans)
refR_2 <- lsmeans(res@objects[[2]], specs = c("dg_cie_10","origen_ingreso_mod"))
ref_dfR_2 <- as.data.frame(cbind(summary(refR_2)[1:2],exp(summary(refR_2)[,c(3,6,7)])))
g4R_2 <- ggplot(ref_dfR_2, aes(x=dg_cie_10, y=lsmean,group=origen_ingreso_mod, colour=origen_ingreso_mod))+
geom_errorbar(aes(ymin=asymp.LCL, ymax=asymp.UCL), width=.1,position=position_dodge(0.1), size=1) +
geom_line(position=position_dodge(0.1), size=.5)+
geom_point(position=position_dodge(0.1), size=2)+
xlab("Psychiatric Condition")+
ylab("Odds Ratio (OR) of Early Drop-out")+
sjPlot::theme_sjplot2() +
geom_rect_interactive(alpha = 0.1, xmin=.1, xmax=.1, ymin=.1,ymax=.1) +
# Remove plot elements added by geom_rect_interactive
theme(legend.position="bottom")+
guides(color=guide_legend(ncol=4,name = "Cause of Discharge"))+
labs(color="Motive of Admission to Treatment")+
scale_colour_brewer(palette = "Set1")+#,labels=c("No Dg.", "In study", "Dg."))+
theme(legend.title = element_blank())
#theme(axis.text = element_blank(), axis.ticks = element_blank())
ggiraph(code = print(g4R_2))library(lsmeans)
refR_3 <- lsmeans(res@objects[[2]], specs = c("dg_cie_10","dg_trs_fisico"))
ref_dfR_3 <- as.data.frame(cbind(summary(refR_3)[1:2],exp(summary(refR_3)[,c(3,6,7)])))
g4R_3 <- ggplot(ref_dfR_3, aes(x=dg_cie_10, y=lsmean,group=dg_trs_fisico, colour=dg_trs_fisico))+
geom_errorbar(aes(ymin=asymp.LCL, ymax=asymp.UCL), width=.1,position=position_dodge(0.1), size=1) +
geom_line(position=position_dodge(0.1), size=.5)+
geom_point(position=position_dodge(0.1), size=2)+
xlab("Physical Condition")+
ylab("Odds Ratio (OR) of Early Drop-out")+
sjPlot::theme_sjplot2() +
geom_rect_interactive(alpha = 0.1, xmin=.1, xmax=.1, ymin=.1,ymax=.1) +
# Remove plot elements added by geom_rect_interactive
theme(legend.position="bottom")+
guides(color=guide_legend(ncol=4,name = "Cause of Discharge"))+
labs(color="Motive of Admission to Treatment")+
scale_colour_brewer(palette = "Set1",labels=c("No Dg.", "In study", "Dg."))+
theme(legend.title = element_blank())
#theme(axis.text = element_blank(), axis.ticks = element_blank())
ggiraph(code = print(g4R_3))save.image("G:/Mi unidad/Alvacast/SISTRAT 2019 (github)/biopsychosoc_comp.RData")
sessionInfo()R version 4.0.2 (2020-06-22)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18363)
Matrix products: default
locale:
[1] LC_COLLATE=Spanish_Chile.1252 LC_CTYPE=Spanish_Chile.1252
[3] LC_MONETARY=Spanish_Chile.1252 LC_NUMERIC=C
[5] LC_TIME=Spanish_Chile.1252
attached base packages:
[1] parallel stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] gdtools_0.2.2 glmulti_1.0.8 leaps_3.1
[4] rJava_0.9-12 reshape2_1.4.4 polycor_0.7-10
[7] DiagrammeR_1.0.6.1.9000 MVN_5.8 citr_0.3.2
[10] ggcorrplot_0.1.3 psych_2.0.7 ggiraph_0.7.0
[13] finalfit_1.0.1 lsmeans_2.30-0 emmeans_1.4.7
[16] choroplethrAdmin1_1.1.1 choroplethrMaps_1.0.1 choroplethr_3.6.3
[19] acs_2.1.4 XML_3.99-0.3 RColorBrewer_1.1-2
[22] panelr_0.7.3 lme4_1.1-23 Matrix_1.2-18
[25] dplyr_1.0.0 data.table_1.12.8 codebook_0.9.2
[28] Statamarkdown_0.4.5 devtools_2.3.0 usethis_1.6.1
[31] sqldf_0.4-11 RSQLite_2.2.0 gsubfn_0.7
[34] proto_1.0.0 broom_0.7.0 zoo_1.8-8
[37] altair_4.0.1 rbokeh_0.5.1 janitor_2.0.1
[40] plotly_4.9.2.1 kableExtra_1.1.0 Hmisc_4.4-0
[43] Formula_1.2-3 survival_3.1-12 lattice_0.20-41
[46] ggplot2_3.3.1 stringr_1.4.0 stringi_1.4.6
[49] tidyr_1.1.0 knitr_1.29 matrixStats_0.56.0
[52] boot_1.3-25
loaded via a namespace (and not attached):
[1] vcd_1.4-7 class_7.3-17 ps_1.3.3
[4] lmtest_0.9-37 rprojroot_1.3-2 crayon_1.3.4
[7] laeken_0.5.1 MASS_7.3-51.6 nlme_3.1-148
[10] backports_1.1.8 rlang_0.4.6 readxl_1.3.1
[13] performance_0.4.6 nloptr_1.2.2.1 callr_3.4.3
[16] rjson_0.2.20 ggmap_3.0.0 bit64_0.9-7
[19] glue_1.4.1 sjPlot_2.8.4 processx_3.4.2
[22] classInt_0.4-3 tcltk_4.0.2 haven_2.3.1
[25] tidyselect_1.1.0 NADA_1.6-1.1 rio_0.5.16
[28] sf_0.9-3 sjmisc_2.8.5 chron_2.3-55
[31] xtable_1.8-4 magrittr_1.5 evaluate_0.14
[34] RgoogleMaps_1.4.5.3 cli_2.0.2 rstudioapi_0.11
[37] miniUI_0.1.1.1 sp_1.4-2 robCompositions_2.2.1
[40] rpart_4.1-15 jtools_2.0.5 pls_2.7-2
[43] zCompositions_1.3.4 sjlabelled_1.1.5 RJSONIO_1.3-1.4
[46] maps_3.3.0 shiny_1.5.0 gistr_0.5.0
[49] xfun_0.16 parameters_0.7.0 pkgbuild_1.1.0
[52] cluster_2.1.0 sgeostat_1.0-27 tibble_3.0.1
[55] permute_0.9-5 png_0.1-7 reshape_0.8.8
[58] withr_2.2.0 bitops_1.0-6 ranger_0.12.1
[61] plyr_1.8.6 cellranger_1.1.0 pcaPP_1.9-73
[64] e1071_1.7-3 coda_0.19-3 pillar_1.4.6
[67] mvoutlier_2.0.9 multcomp_1.4-13 fs_1.4.1
[70] flexmix_2.3-15 kernlab_0.9-29 vctrs_0.3.1
[73] ellipsis_0.3.1 generics_0.0.2 nortest_1.0-4
[76] rgdal_1.5-8 tools_4.0.2 foreign_0.8-80
[79] munsell_0.5.0 fastmap_1.0.1 compiler_4.0.2
[82] pkgload_1.1.0 abind_1.4-5 httpuv_1.5.4
[85] tigris_0.9.4 npmv_2.4.0 sessioninfo_1.1.1
[88] gridExtra_2.3 visNetwork_2.0.9 later_1.1.0.1
[91] jsonlite_1.6.1 WDI_2.6.0 GGally_1.5.0
[94] scales_1.1.1 carData_3.0-4 estimability_1.3
[97] lazyeval_0.2.2 promises_1.1.1 car_3.0-8
[100] latticeExtra_0.6-29 reticulate_1.16 effectsize_0.3.1
[103] checkmate_2.0.0 rmarkdown_2.3 openxlsx_4.1.5
[106] sandwich_2.5-1 moments_0.14 statmod_1.4.34
[109] webshot_0.5.2 forcats_0.5.0 pander_0.6.3
[112] yaml_2.2.1 systemfonts_0.2.3 prabclus_2.3-2
[115] htmltools_0.5.0 memoise_1.1.0 modeltools_0.2-23
[118] viridisLite_0.3.0 digest_0.6.25 rrcov_1.5-2
[121] assertthat_0.2.1 mime_0.9 rappdirs_0.3.1
[124] repr_1.1.0 bayestestR_0.6.0 units_0.6-6
[127] remotes_2.1.1 vegan_2.5-6 blob_1.2.1
[130] splines_4.0.2 labeling_0.3 fpc_2.2-7
[133] cvTools_0.3.2 hms_0.5.3 modelr_0.1.8
[136] colorspace_1.4-1 base64enc_0.1-3 mnormt_2.0.0
[139] tmvnsim_1.0-2 nnet_7.3-14 Rcpp_1.0.4.6
[142] mclust_5.4.6 mvtnorm_1.1-1 fansi_0.4.1
[145] VIM_6.0.0 truncnorm_1.0-8 R6_2.4.1
[148] grid_4.0.2 lifecycle_0.2.0 acepack_1.4.1
[151] labelled_2.5.0 zip_2.0.4 curl_4.3
[154] pryr_0.1.4 minqa_1.2.4 testthat_2.3.2
[157] snakecase_0.11.0 robustbase_0.93-6 skimr_2.1.1
[160] desc_1.2.0 TH.data_1.0-10 htmlwidgets_1.5.1
[163] purrr_0.3.4 crosstalk_1.1.0.1 mgcv_1.8-31
[166] rvest_0.3.5 insight_0.8.4 htmlTable_1.13.3
[169] codetools_0.2-16 lubridate_1.7.9 prettyunits_1.1.1
[172] vegawidget_0.3.1 gtable_0.3.0 DBI_1.1.0
[175] stats4_4.0.2 httr_1.4.2 highr_0.8
[178] KernSmooth_2.23-17 farver_2.0.3 uuid_0.1-4
[181] diptest_0.75-7 hexbin_1.28.1 mice_3.9.0
[184] xml2_1.3.2 ggeffects_0.14.3 readr_1.3.1
[187] sROC_0.1-2 energy_1.7-7 DEoptimR_1.0-8
[190] bit_1.1-15.2 sjstats_0.18.0 jpeg_0.1-8.1
[193] pkgconfig_2.0.3 maptools_1.0-1 References
1. Revelle W. psych: Procedures for personality and psychological research.
2. Fox J. polycor: Polychoric and Polyserial Correlations.
3. Silva J da, Ventura CAA, Vargens OM da C, et al. Illicit drug use in seven Latin American countries: critical perspectives of families and familiars. Revista Latino-Americana de Enfermagem 2009; 17: 763–769.
4. Basu D, Ghosh A, Sarkar S, et al. Initial treatment dropout in patients with substance use disorders attending a tertiary care de-addiction centre in north India. The Indian journal of medical research 2017; 146: S77–S84.
5. Lappan SN, Brown AW, Hendricks PS. Dropout rates of in-person psychosocial substance use disorder treatments: a systematic review and meta-analysis. Addiction 2020; 115: 201–217.
6. Pérez-Gómez A, Mejía-Trujillo J. The Evolution of Alcohol and Drug Prevention Strategies in Latin America. In: Romano JL, Israelashvili M (eds) The cambridge handbook of international prevention science. Cambridge: Cambridge University Press, pp. 753–779.
7. Ronzani TM, Fuentes-Mejía C, Mota DCB, et al. Intervenciones breves para el abuso de sustancias en América Latina: una revisión sistemática. Psicologia em Estudo; 24, http://www.scielo.br/scielo.php?script=sci{\_}arttext{\&}pid=S1413-73722019000100232{\&}nrm=iso (2019).
8. Heinze G, Armas-Castañeda G. Public policies on the use of drugs in Mexico and Latin America. Drug Science, Policy and Law 2015; 2: 2050324515611587.
9. Calcagno V, Mazancourt C de. glmulti: an R package for easy automated model selection with (generalized) linear models. Epub ahead of print 2010. DOI: 10.18637/jss.v034.i12.
10. Dixon P. VEGAN, a package of R functions for community ecology. Journal of Vegetation Science 2003; 14: 927–930.
11. Team RC. R: A language and environment for statistical computing., http://www.r-project.org/ (2020).
12. Inter-American Drug Abuse Control Commission [CICAD]. Report on Drug Use in the Americas 2019. 2019. OEA/Ser.L/XIV.6.6, Washington, D.C.: Organization of American States (OAS), http://www.cicad.oas.org/main/pubs/Report on Drug Use in the Americas 2019.pdf (2019).
13. Korkmaz D. S, Zararsiz G. MVN: An R Package for Assessing Multivariate Normality. The R Journal 2014; 6: 151–162.
14. Burchett W, Ellis A, Harrar S, et al. Nonparametric Inference for Multivariate Data: The R Package npmv. Journal of Statistical Software; 76. Epub ahead of print 2017. DOI: 10.18637/jss.v076.i04.
15. Soto-Brandt G, Portilla Huidobro R, Huepe Artigas D, et al. [Validity evidence of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) in Chile]. Adicciones 2014; 26: 291–302.
16. Hansen Z. A Phenomenological Investigation of Clinical Intuition among Alcohol and Drug Counselors. 2015.
17. Stone J, Marsh A, Dale A, et al. Counselling Guideleines: Fourth Edition: Alchol and Other Drug Issues. Western Australian Government - Mental Health Commission, https://books.google.cl/books?id=K{\_}k2xQEACAAJ (2019).
18. Group CG on DM, Working DU2IE. Drug misuse and dependence: UK guidelines on clinical management. Public Health England. Alcohol, Drugs & Tobacco Division, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment{\_}data/file/673978/clinical{\_}guidelines{\_}2017.pdf (2017).
19. Center for Substance Abuse T. SAMHSA/CSAT Treatment Improvement Protocols. In: Substance abuse treatment for persons with co-occurring disorders. Rockville (MD): Substance Abuse; Mental Health Services Administration (US), 2005.
20. Maunder RG, Wiesenfeld L, Lawson A, et al. The Relationship Between Childhood Adversity and Other Aspects of Clinical Complexity in Psychiatric Outpatients. Journal of Interpersonal Violence 2019; 0886260519865968.
21. Hsieh M-H, Tsai S-L, Tsai C-H, et al. What Is the Addiction World Like? Understanding the Lived Experience of the Individuals’ Illicit Drug Addiction in Taiwan. Perspectives in Psychiatric Care 2017; 53: 47–54.
22. Messas G, Fukuda L, Pienkos E. A Phenomenological Contribution to Substance Misuse Treatment: Principles for Person-Centered Care. Psychopathology 2019; 52: 1–9.
23. Copoeru I. Understanding Addiction: A Threefold Phenomenological Approach. Human Studies 2014; 37: 335–349.
24. Pachado MP, Scherer JN, Guimarães LSP, et al. Markers for Severity of Problems in Interpersonal Relationships of Crack Cocaine Users from a Brazilian Multicenter Study. Psychiatric Quarterly 2018; 89: 923–936.
25. Ministerio de Salud[MINSAL]. Norma y orientaciones técnicas de los planes de tratamiento y rehabilitación para personas adultas con problemas derivados del consumo de drogas. 391, https://www.senda.gob.cl/wp-content/uploads/2012/08/OrientacionesTecnicas{\_}CentrosdeTratamiento.pdf (2012).
26. Ministerio de Salud[MINSAL]. Norma técnica para el tratamiento integral de adolescentes infractores de ley con consumo problemático de alcohol - drogas y otros trastornos de salud mental. 391, https://www.minsal.cl/portal/url/item/71e5abf67b425395e04001011f017d2e.pdf (2006).
27. Valdivieso M. Sistematización de las prácticas de un programa de rehabilitación en drogas y alcohol para mujeres, con enfoque de género, que ha venido realizando el equipo de profesionales y técnicos del CTR del hospital de Peñablanca. PhD thesis, Universidad de Chile, http://repositorio.uchile.cl/handle/2250/135697 (2014).
28. Consejo Nacional para el Control de Estupefacientes[CONACE]. Modelo de intervención en personas con consumo problemático de sustancias psicoactivas, recluidas en los establecimientos penitenciarios chilenos. Tomo II. Programa de Tratamiento, Rehabilitación y Reinserción Social, para internos/as con consumo problemático de sustancias psicoactivas. Ministerio del Interior, http://sistemas.senda.gob.cl/sistema-monitoreo/biblioteca/files/Documentos/ESTRATEGIAS NORMAS ORIENTACIONES/1 Orientaciones y Normas/Nacional/Senda/Modelo de intervenci{\'{o}}n de personas recluidas en establecimientos penitenciarios VOL 2.pdf (2005).
29. Consejo Nacional para el Control de Estupefacientes [CONACE]. Orientaciones Técnicas. Tratamiento del consumo problemático de alcohol y drogas y otros trastornos de salud mental en adolescentes infractores de ley. 2007.
30. Consejo Nacional para el Control de Estupefacientes[CONACE]. Modelo de intervención en personas con consumo problemático de sustancias psicoactivas, recluidas en los establecimientos penitenciarios chilenos. Tomo I. Elementos Teóricos del Programa de Tratamiento, Rehabilitación y Reinserción Social, para Internos/as con consumo problemático de sustancias psicoactivas. Ministerio del Interior, http://sistemas.senda.gob.cl/sistema-monitoreo/biblioteca/files/Documentos/ESTRATEGIAS NORMAS ORIENTACIONES/1 Orientaciones y Normas/Nacional/Senda/Modelo de intervenci{\'{o}}n de personas recluidas en establecimientos penitenciarios VOL 1.pdf (2005).
31. Huidobro R, Lozier M, Oliva M. Informe resultados: Treatment Outcomes Profile (TOP) Chile. 2016.
32. Tiffany ST, Friedman L, Greenfield SF, et al. Beyond drug use: a systematic consideration of other outcomes in evaluations of treatments for substance use disorders. Addiction (Abingdon, England) 2012; 107: 709–718.
33. National Institute on Drug Abuse[NIDA]. Part 1: The Connection Between Substance Use Disorders and Mental Illness. National Institute on Drug Abuse, https://www.drugabuse.gov/publications/research-reports/common-comorbidities-substance-use-disorders/part-1-connection-between-substance-use-disorders-mental-illness.
34. Simoneau H, Brochu S. Addiction severity index profile of persons who reenter treatment for substance use disorders. Substance Abuse 2017; 38: 432–437.
35. Leung J, Peacock A, Colledge S, et al. A Global Meta-analysis of the Prevalence of HIV, Hepatitis C Virus, and Hepatitis B Virus Among People Who Inject Drugs—Do Gender-Based Differences Vary by Country-Level Indicators? The Journal of Infectious Diseases 2019; 220: 78–90.
36. Lozano ÓM, Rojas AJ, Fernández Calderón F. Psychiatric comorbidity and severity of dependence on substance users: how it impacts on their health-related quality of life? Journal of Mental Health 2017; 26: 119–126.
37. Kopak AM, Combs E, Goodman K, et al. Exposure to Violence and Substance Use Treatment Outcomes Among Female Patients. Substance Use & Misuse 2019; 54: 362–372.
38. Medlock MM, Rosmarin DH, Connery HS, et al. Religious coping in patients with severe substance use disorders receiving acute inpatient detoxification. The American journal on addictions 2017; 26: 744–750.
39. Hildebrandt T, Epstein EE, Sysko R, et al. Using Factor Mixture Models to Evaluate the Type A/B Classification of Alcohol Use Disorders in a Heterogeneous Treatment Sample. Alcoholism, clinical and experimental research 2017; 41: 987–997.
40. Akbari H, Roshanpajouh M, Nourijelyani K, et al. Profile of drug users in the residential treatment centers of Tehran, Iran. Health promotion perspectives 2019; 9: 248–254.
41. Saberi Zafarghandi MB, Khanipour H, Ahmadi SM. Typology of Substance Use Disorder Based on Temperament Dimensions, Addiction Severity, and Negative Emotions. Iranian journal of psychiatry 2018; 13: 184–190.
42. Johannessen DA, Nordfjærn T, Geirdal AØ. Change in psychosocial factors connected to coping after inpatient treatment for substance use disorder: a systematic review. Substance abuse treatment, prevention, and policy 2019; 14: 16.
43. Lopes-Rosa R, Kessler FP, Pianca TG, et al. Predictors of early relapse among adolescent crack users. Journal of Addictive Diseases 2017; 36: 136–143.
44. Klein M. Relapse into opiate and crack cocaine misuse: a scoping review. Addiction Research & Theory 2020; 1–19.
45. Schmid M, Michaud L, Bovio N, et al. Prevalence of somatic and psychiatric morbidity across occupations in Switzerland and its correlation with suicide mortality: results from the Swiss National Cohort (1990–2014). BMC Psychiatry 2020; 20: 324.
46. Gmel G, Akre C, Astudillo M, et al. The Swiss Cohort Study on Substance Use Risk Factors - Findings of two Waves. Sucht 2015; 61: 251–262.
47. Hjemsæter AJ, Bramness JG, Drake R, et al. Mortality, cause of death and risk factors in patients with alcohol use disorder alone or poly-substance use disorders: a 19-year prospective cohort study. BMC Psychiatry 2019; 19: 101.
48. Moradinazar M, Najafi F, Jalilian F, et al. Prevalence of drug use, alcohol consumption, cigarette smoking and measure of socioeconomic-related inequalities of drug use among Iranian people: findings from a national survey. Substance abuse treatment, prevention, and policy 2020; 15: 39.
50. Saberi H.Ahmadi, SM. M. Typology of Substance Use Disorder Based on Temperament Dimensions, Addiction Severity, and Negative Emotions. Iran J Psychiatry 2018; 13: 184–190.
51. Portela A, Garcia L, Goldim J. Adolescencia vulnerable: factores biopsicosociales relacionados al uso de drogas. Revista Bioética 2015; 23: 311–319.
52. Papadimitriou G. The "Biopsychosocial Model": 40 years of application in Psychiatry. Psychiatriki 2017; 28: 107–110.
53. Organization[WHO] WH. World Drug Report 2019 (Sales No. E.19.XI.8). Viena, 2020.
54. Teesson M, Degenhardt L, Hall W. Theories of addiction: Causes and maintenance of addiction. In: Addictions. London: Psychology Press, 2002, pp. 33–47.
55. Vos T, Allen C, Arora M, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet. Epub ahead of print 2016. DOI: 10.1016/S0140-6736(16)31678-6.
56. Zijlmans EAO, Ark LA van der, Tijmstra J, et al. Methods for Estimating Item-Score Reliability. Applied Psychological Measurement. Epub ahead of print 2018. DOI: 10.1177/0146621618758290.
57. Castillo-Carniglia Á, Marín JD, Soto-Brandt G, et al. Adaptation and Validation of the Instrument Treatment Outcomes Profile to the Chilean Population. Journal of Substance Abuse Treatment. Epub ahead of print 2015. DOI: 10.1016/j.jsat.2015.03.002.
58. Collaborators GBD2D, Incidence I, Prevalence. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet (London, England) 2018; 392: 1789–1858.
59. Polychoric T, Correlations P, Fox AJ. Package ‘polycor’. Statistics.
60. Pacurucu-Castillo SF, Ordóñez-Mancheno JM, Hernández-Cruz A, et al. World Opioid and Substance Use Epidemic: A Latin American Perspective. Psychiatric Research and Clinical Practice. Epub ahead of print 2019. DOI: 10.1176/appi.prcp.20180009.
61. De Boni RB, Peratikos MB, Shepherd BE, et al. Is substance use associated with HIV cascade outcomes in Latin America? PLoS ONE. Epub ahead of print 2018. DOI: 10.1371/journal.pone.0194228.